ThetaRay’s IntuitiveAI solutions replicate the powerful decision-making capabilities of human intuition to detect “unknown unknowns” that could not be identified by first-generation AI or legacy products.
ThetaRay’s machine learning algorithms that comprise IntuitiveAI were developed by two world-renowned mathematicians. They analyze massive amounts of data and discover relationships between seemingly unrelated events. They enable banks to pinpoint activity that suggests money laundering, terrorist financing, human and drug trafficking, and other financial crimes.
Version 4.0 provides a new hybrid learning approach. The hybrid supervised/unsupervised learning capability integrates the two learning styles and applies the most effective one based on use case. This approach finds significantly more potential threats through a single process and delivers a holistic view of a bank’s threat landscape.
The new release also provides an additional method for anomaly clustering, which is a critical enabler to accurately detect more true positives while dramatically decreasing the number of false positive alerts.
In version 4.0, customers can now cluster identified anomalies by pattern, in addition to a density-clustering approach. This clustering method ensures that AML and fraud teams have the right approach to analyze anomalous events with the method most applicable to a particular use case. The addition of pattern-based clustering also enhances the built-in transparency and explainability of ThetaRay’s “white box” AI applications.
ThetaRay’s enterprise services foundation has been enhanced with additional API capabilities to make it easier for DevOps teams to automate financial crime and related solutions. They also make ThetaRay’s system even easier to deploy, seamlessly integrating into existing workflows and processes so that businesses can continue leveraging their investments in existing infrastructure. ThetaRay 4.0 also advances the company’s cloud strategy with HDFS support on Microsoft Azure.
ThetaRay - www.thetaray.com